File size: 24,257 Bytes
d8f1638 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 |
2023-10-23 14:51:33,665 ----------------------------------------------------------------------------------------------------
2023-10-23 14:51:33,666 Model: "SequenceTagger(
(embeddings): TransformerWordEmbeddings(
(model): BertModel(
(embeddings): BertEmbeddings(
(word_embeddings): Embedding(64001, 768)
(position_embeddings): Embedding(512, 768)
(token_type_embeddings): Embedding(2, 768)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(encoder): BertEncoder(
(layer): ModuleList(
(0-11): 12 x BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
)
(pooler): BertPooler(
(dense): Linear(in_features=768, out_features=768, bias=True)
(activation): Tanh()
)
)
)
(locked_dropout): LockedDropout(p=0.5)
(linear): Linear(in_features=768, out_features=25, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-23 14:51:33,666 ----------------------------------------------------------------------------------------------------
2023-10-23 14:51:33,666 MultiCorpus: 1100 train + 206 dev + 240 test sentences
- NER_HIPE_2022 Corpus: 1100 train + 206 dev + 240 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/ajmc/de/with_doc_seperator
2023-10-23 14:51:33,666 ----------------------------------------------------------------------------------------------------
2023-10-23 14:51:33,666 Train: 1100 sentences
2023-10-23 14:51:33,666 (train_with_dev=False, train_with_test=False)
2023-10-23 14:51:33,666 ----------------------------------------------------------------------------------------------------
2023-10-23 14:51:33,666 Training Params:
2023-10-23 14:51:33,666 - learning_rate: "3e-05"
2023-10-23 14:51:33,666 - mini_batch_size: "4"
2023-10-23 14:51:33,666 - max_epochs: "10"
2023-10-23 14:51:33,666 - shuffle: "True"
2023-10-23 14:51:33,666 ----------------------------------------------------------------------------------------------------
2023-10-23 14:51:33,666 Plugins:
2023-10-23 14:51:33,666 - TensorboardLogger
2023-10-23 14:51:33,666 - LinearScheduler | warmup_fraction: '0.1'
2023-10-23 14:51:33,666 ----------------------------------------------------------------------------------------------------
2023-10-23 14:51:33,667 Final evaluation on model from best epoch (best-model.pt)
2023-10-23 14:51:33,667 - metric: "('micro avg', 'f1-score')"
2023-10-23 14:51:33,667 ----------------------------------------------------------------------------------------------------
2023-10-23 14:51:33,667 Computation:
2023-10-23 14:51:33,667 - compute on device: cuda:0
2023-10-23 14:51:33,667 - embedding storage: none
2023-10-23 14:51:33,667 ----------------------------------------------------------------------------------------------------
2023-10-23 14:51:33,667 Model training base path: "hmbench-ajmc/de-dbmdz/bert-base-historic-multilingual-64k-td-cased-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-1"
2023-10-23 14:51:33,667 ----------------------------------------------------------------------------------------------------
2023-10-23 14:51:33,667 ----------------------------------------------------------------------------------------------------
2023-10-23 14:51:33,667 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-23 14:51:36,118 epoch 1 - iter 27/275 - loss 3.76030068 - time (sec): 2.45 - samples/sec: 920.61 - lr: 0.000003 - momentum: 0.000000
2023-10-23 14:51:37,502 epoch 1 - iter 54/275 - loss 2.96902549 - time (sec): 3.83 - samples/sec: 1141.80 - lr: 0.000006 - momentum: 0.000000
2023-10-23 14:51:38,999 epoch 1 - iter 81/275 - loss 2.35139223 - time (sec): 5.33 - samples/sec: 1198.48 - lr: 0.000009 - momentum: 0.000000
2023-10-23 14:51:40,377 epoch 1 - iter 108/275 - loss 1.89822180 - time (sec): 6.71 - samples/sec: 1282.13 - lr: 0.000012 - momentum: 0.000000
2023-10-23 14:51:41,879 epoch 1 - iter 135/275 - loss 1.64164622 - time (sec): 8.21 - samples/sec: 1327.52 - lr: 0.000015 - momentum: 0.000000
2023-10-23 14:51:43,282 epoch 1 - iter 162/275 - loss 1.44032124 - time (sec): 9.61 - samples/sec: 1354.11 - lr: 0.000018 - momentum: 0.000000
2023-10-23 14:51:44,700 epoch 1 - iter 189/275 - loss 1.28091475 - time (sec): 11.03 - samples/sec: 1384.41 - lr: 0.000021 - momentum: 0.000000
2023-10-23 14:51:46,155 epoch 1 - iter 216/275 - loss 1.14720675 - time (sec): 12.49 - samples/sec: 1410.01 - lr: 0.000023 - momentum: 0.000000
2023-10-23 14:51:47,580 epoch 1 - iter 243/275 - loss 1.03078045 - time (sec): 13.91 - samples/sec: 1453.56 - lr: 0.000026 - momentum: 0.000000
2023-10-23 14:51:48,990 epoch 1 - iter 270/275 - loss 0.95387520 - time (sec): 15.32 - samples/sec: 1455.89 - lr: 0.000029 - momentum: 0.000000
2023-10-23 14:51:49,295 ----------------------------------------------------------------------------------------------------
2023-10-23 14:51:49,295 EPOCH 1 done: loss 0.9430 - lr: 0.000029
2023-10-23 14:51:49,715 DEV : loss 0.19450944662094116 - f1-score (micro avg) 0.754
2023-10-23 14:51:49,720 saving best model
2023-10-23 14:51:50,181 ----------------------------------------------------------------------------------------------------
2023-10-23 14:51:51,576 epoch 2 - iter 27/275 - loss 0.16817983 - time (sec): 1.39 - samples/sec: 1484.89 - lr: 0.000030 - momentum: 0.000000
2023-10-23 14:51:52,989 epoch 2 - iter 54/275 - loss 0.13521295 - time (sec): 2.81 - samples/sec: 1555.17 - lr: 0.000029 - momentum: 0.000000
2023-10-23 14:51:54,461 epoch 2 - iter 81/275 - loss 0.15603419 - time (sec): 4.28 - samples/sec: 1534.63 - lr: 0.000029 - momentum: 0.000000
2023-10-23 14:51:55,873 epoch 2 - iter 108/275 - loss 0.16617353 - time (sec): 5.69 - samples/sec: 1551.22 - lr: 0.000029 - momentum: 0.000000
2023-10-23 14:51:57,276 epoch 2 - iter 135/275 - loss 0.16037825 - time (sec): 7.09 - samples/sec: 1569.62 - lr: 0.000028 - momentum: 0.000000
2023-10-23 14:51:58,579 epoch 2 - iter 162/275 - loss 0.16018872 - time (sec): 8.40 - samples/sec: 1587.20 - lr: 0.000028 - momentum: 0.000000
2023-10-23 14:51:59,976 epoch 2 - iter 189/275 - loss 0.16213256 - time (sec): 9.79 - samples/sec: 1603.60 - lr: 0.000028 - momentum: 0.000000
2023-10-23 14:52:01,268 epoch 2 - iter 216/275 - loss 0.16255960 - time (sec): 11.09 - samples/sec: 1601.19 - lr: 0.000027 - momentum: 0.000000
2023-10-23 14:52:02,575 epoch 2 - iter 243/275 - loss 0.16177479 - time (sec): 12.39 - samples/sec: 1626.33 - lr: 0.000027 - momentum: 0.000000
2023-10-23 14:52:03,884 epoch 2 - iter 270/275 - loss 0.16005825 - time (sec): 13.70 - samples/sec: 1637.61 - lr: 0.000027 - momentum: 0.000000
2023-10-23 14:52:04,167 ----------------------------------------------------------------------------------------------------
2023-10-23 14:52:04,167 EPOCH 2 done: loss 0.1590 - lr: 0.000027
2023-10-23 14:52:04,693 DEV : loss 0.14087921380996704 - f1-score (micro avg) 0.8186
2023-10-23 14:52:04,698 saving best model
2023-10-23 14:52:05,312 ----------------------------------------------------------------------------------------------------
2023-10-23 14:52:06,704 epoch 3 - iter 27/275 - loss 0.08749511 - time (sec): 1.39 - samples/sec: 1529.91 - lr: 0.000026 - momentum: 0.000000
2023-10-23 14:52:08,163 epoch 3 - iter 54/275 - loss 0.08996595 - time (sec): 2.85 - samples/sec: 1567.37 - lr: 0.000026 - momentum: 0.000000
2023-10-23 14:52:09,565 epoch 3 - iter 81/275 - loss 0.07783644 - time (sec): 4.25 - samples/sec: 1647.07 - lr: 0.000026 - momentum: 0.000000
2023-10-23 14:52:10,958 epoch 3 - iter 108/275 - loss 0.08738457 - time (sec): 5.64 - samples/sec: 1587.28 - lr: 0.000025 - momentum: 0.000000
2023-10-23 14:52:12,556 epoch 3 - iter 135/275 - loss 0.09031569 - time (sec): 7.24 - samples/sec: 1565.46 - lr: 0.000025 - momentum: 0.000000
2023-10-23 14:52:13,940 epoch 3 - iter 162/275 - loss 0.08700344 - time (sec): 8.63 - samples/sec: 1551.18 - lr: 0.000025 - momentum: 0.000000
2023-10-23 14:52:15,368 epoch 3 - iter 189/275 - loss 0.09749654 - time (sec): 10.06 - samples/sec: 1544.21 - lr: 0.000024 - momentum: 0.000000
2023-10-23 14:52:16,673 epoch 3 - iter 216/275 - loss 0.09678369 - time (sec): 11.36 - samples/sec: 1557.87 - lr: 0.000024 - momentum: 0.000000
2023-10-23 14:52:17,998 epoch 3 - iter 243/275 - loss 0.09737047 - time (sec): 12.68 - samples/sec: 1579.72 - lr: 0.000024 - momentum: 0.000000
2023-10-23 14:52:19,404 epoch 3 - iter 270/275 - loss 0.09843592 - time (sec): 14.09 - samples/sec: 1586.75 - lr: 0.000023 - momentum: 0.000000
2023-10-23 14:52:19,664 ----------------------------------------------------------------------------------------------------
2023-10-23 14:52:19,664 EPOCH 3 done: loss 0.0972 - lr: 0.000023
2023-10-23 14:52:20,208 DEV : loss 0.16271711885929108 - f1-score (micro avg) 0.832
2023-10-23 14:52:20,213 saving best model
2023-10-23 14:52:20,764 ----------------------------------------------------------------------------------------------------
2023-10-23 14:52:22,164 epoch 4 - iter 27/275 - loss 0.09045084 - time (sec): 1.40 - samples/sec: 1553.16 - lr: 0.000023 - momentum: 0.000000
2023-10-23 14:52:23,595 epoch 4 - iter 54/275 - loss 0.08002820 - time (sec): 2.83 - samples/sec: 1622.98 - lr: 0.000023 - momentum: 0.000000
2023-10-23 14:52:24,901 epoch 4 - iter 81/275 - loss 0.07686002 - time (sec): 4.14 - samples/sec: 1628.96 - lr: 0.000022 - momentum: 0.000000
2023-10-23 14:52:26,261 epoch 4 - iter 108/275 - loss 0.07765183 - time (sec): 5.50 - samples/sec: 1624.85 - lr: 0.000022 - momentum: 0.000000
2023-10-23 14:52:27,647 epoch 4 - iter 135/275 - loss 0.07194057 - time (sec): 6.88 - samples/sec: 1615.78 - lr: 0.000022 - momentum: 0.000000
2023-10-23 14:52:29,032 epoch 4 - iter 162/275 - loss 0.07075261 - time (sec): 8.27 - samples/sec: 1577.70 - lr: 0.000021 - momentum: 0.000000
2023-10-23 14:52:30,432 epoch 4 - iter 189/275 - loss 0.07211105 - time (sec): 9.67 - samples/sec: 1587.44 - lr: 0.000021 - momentum: 0.000000
2023-10-23 14:52:31,797 epoch 4 - iter 216/275 - loss 0.07529380 - time (sec): 11.03 - samples/sec: 1613.99 - lr: 0.000021 - momentum: 0.000000
2023-10-23 14:52:33,196 epoch 4 - iter 243/275 - loss 0.07198357 - time (sec): 12.43 - samples/sec: 1609.00 - lr: 0.000020 - momentum: 0.000000
2023-10-23 14:52:34,569 epoch 4 - iter 270/275 - loss 0.07106318 - time (sec): 13.80 - samples/sec: 1621.75 - lr: 0.000020 - momentum: 0.000000
2023-10-23 14:52:34,812 ----------------------------------------------------------------------------------------------------
2023-10-23 14:52:34,812 EPOCH 4 done: loss 0.0712 - lr: 0.000020
2023-10-23 14:52:35,342 DEV : loss 0.156390979886055 - f1-score (micro avg) 0.8345
2023-10-23 14:52:35,347 saving best model
2023-10-23 14:52:35,958 ----------------------------------------------------------------------------------------------------
2023-10-23 14:52:37,327 epoch 5 - iter 27/275 - loss 0.03811944 - time (sec): 1.37 - samples/sec: 1511.28 - lr: 0.000020 - momentum: 0.000000
2023-10-23 14:52:38,717 epoch 5 - iter 54/275 - loss 0.04905995 - time (sec): 2.76 - samples/sec: 1566.02 - lr: 0.000019 - momentum: 0.000000
2023-10-23 14:52:40,114 epoch 5 - iter 81/275 - loss 0.04532731 - time (sec): 4.15 - samples/sec: 1582.81 - lr: 0.000019 - momentum: 0.000000
2023-10-23 14:52:41,530 epoch 5 - iter 108/275 - loss 0.05491363 - time (sec): 5.57 - samples/sec: 1567.00 - lr: 0.000019 - momentum: 0.000000
2023-10-23 14:52:42,913 epoch 5 - iter 135/275 - loss 0.05232614 - time (sec): 6.95 - samples/sec: 1563.41 - lr: 0.000018 - momentum: 0.000000
2023-10-23 14:52:44,364 epoch 5 - iter 162/275 - loss 0.04940998 - time (sec): 8.40 - samples/sec: 1565.20 - lr: 0.000018 - momentum: 0.000000
2023-10-23 14:52:45,749 epoch 5 - iter 189/275 - loss 0.04778139 - time (sec): 9.79 - samples/sec: 1578.83 - lr: 0.000018 - momentum: 0.000000
2023-10-23 14:52:47,153 epoch 5 - iter 216/275 - loss 0.05341423 - time (sec): 11.19 - samples/sec: 1585.88 - lr: 0.000017 - momentum: 0.000000
2023-10-23 14:52:48,596 epoch 5 - iter 243/275 - loss 0.05397692 - time (sec): 12.64 - samples/sec: 1586.48 - lr: 0.000017 - momentum: 0.000000
2023-10-23 14:52:49,977 epoch 5 - iter 270/275 - loss 0.05183480 - time (sec): 14.02 - samples/sec: 1594.84 - lr: 0.000017 - momentum: 0.000000
2023-10-23 14:52:50,228 ----------------------------------------------------------------------------------------------------
2023-10-23 14:52:50,228 EPOCH 5 done: loss 0.0517 - lr: 0.000017
2023-10-23 14:52:50,767 DEV : loss 0.1555197536945343 - f1-score (micro avg) 0.8662
2023-10-23 14:52:50,772 saving best model
2023-10-23 14:52:51,322 ----------------------------------------------------------------------------------------------------
2023-10-23 14:52:52,768 epoch 6 - iter 27/275 - loss 0.05170272 - time (sec): 1.44 - samples/sec: 1613.70 - lr: 0.000016 - momentum: 0.000000
2023-10-23 14:52:54,152 epoch 6 - iter 54/275 - loss 0.03843071 - time (sec): 2.83 - samples/sec: 1584.45 - lr: 0.000016 - momentum: 0.000000
2023-10-23 14:52:55,541 epoch 6 - iter 81/275 - loss 0.03440502 - time (sec): 4.22 - samples/sec: 1615.84 - lr: 0.000016 - momentum: 0.000000
2023-10-23 14:52:56,991 epoch 6 - iter 108/275 - loss 0.03812148 - time (sec): 5.67 - samples/sec: 1616.93 - lr: 0.000015 - momentum: 0.000000
2023-10-23 14:52:58,368 epoch 6 - iter 135/275 - loss 0.03963989 - time (sec): 7.04 - samples/sec: 1608.15 - lr: 0.000015 - momentum: 0.000000
2023-10-23 14:52:59,844 epoch 6 - iter 162/275 - loss 0.04065051 - time (sec): 8.52 - samples/sec: 1586.96 - lr: 0.000015 - momentum: 0.000000
2023-10-23 14:53:01,230 epoch 6 - iter 189/275 - loss 0.03839786 - time (sec): 9.91 - samples/sec: 1594.46 - lr: 0.000014 - momentum: 0.000000
2023-10-23 14:53:02,632 epoch 6 - iter 216/275 - loss 0.04073106 - time (sec): 11.31 - samples/sec: 1589.21 - lr: 0.000014 - momentum: 0.000000
2023-10-23 14:53:04,056 epoch 6 - iter 243/275 - loss 0.03969929 - time (sec): 12.73 - samples/sec: 1591.47 - lr: 0.000014 - momentum: 0.000000
2023-10-23 14:53:05,456 epoch 6 - iter 270/275 - loss 0.04061644 - time (sec): 14.13 - samples/sec: 1581.07 - lr: 0.000013 - momentum: 0.000000
2023-10-23 14:53:05,717 ----------------------------------------------------------------------------------------------------
2023-10-23 14:53:05,717 EPOCH 6 done: loss 0.0402 - lr: 0.000013
2023-10-23 14:53:06,253 DEV : loss 0.16560792922973633 - f1-score (micro avg) 0.8671
2023-10-23 14:53:06,258 saving best model
2023-10-23 14:53:06,859 ----------------------------------------------------------------------------------------------------
2023-10-23 14:53:08,317 epoch 7 - iter 27/275 - loss 0.00265191 - time (sec): 1.46 - samples/sec: 1397.17 - lr: 0.000013 - momentum: 0.000000
2023-10-23 14:53:09,652 epoch 7 - iter 54/275 - loss 0.01398723 - time (sec): 2.79 - samples/sec: 1563.25 - lr: 0.000013 - momentum: 0.000000
2023-10-23 14:53:11,083 epoch 7 - iter 81/275 - loss 0.02775533 - time (sec): 4.22 - samples/sec: 1556.99 - lr: 0.000012 - momentum: 0.000000
2023-10-23 14:53:12,449 epoch 7 - iter 108/275 - loss 0.03402506 - time (sec): 5.59 - samples/sec: 1606.20 - lr: 0.000012 - momentum: 0.000000
2023-10-23 14:53:13,804 epoch 7 - iter 135/275 - loss 0.02753248 - time (sec): 6.94 - samples/sec: 1624.36 - lr: 0.000012 - momentum: 0.000000
2023-10-23 14:53:15,228 epoch 7 - iter 162/275 - loss 0.03091741 - time (sec): 8.37 - samples/sec: 1636.47 - lr: 0.000011 - momentum: 0.000000
2023-10-23 14:53:16,616 epoch 7 - iter 189/275 - loss 0.02768566 - time (sec): 9.75 - samples/sec: 1627.12 - lr: 0.000011 - momentum: 0.000000
2023-10-23 14:53:17,992 epoch 7 - iter 216/275 - loss 0.02673091 - time (sec): 11.13 - samples/sec: 1619.87 - lr: 0.000011 - momentum: 0.000000
2023-10-23 14:53:19,436 epoch 7 - iter 243/275 - loss 0.02610998 - time (sec): 12.57 - samples/sec: 1601.55 - lr: 0.000010 - momentum: 0.000000
2023-10-23 14:53:20,816 epoch 7 - iter 270/275 - loss 0.02775259 - time (sec): 13.96 - samples/sec: 1602.26 - lr: 0.000010 - momentum: 0.000000
2023-10-23 14:53:21,071 ----------------------------------------------------------------------------------------------------
2023-10-23 14:53:21,071 EPOCH 7 done: loss 0.0280 - lr: 0.000010
2023-10-23 14:53:21,597 DEV : loss 0.16052626073360443 - f1-score (micro avg) 0.875
2023-10-23 14:53:21,603 saving best model
2023-10-23 14:53:22,286 ----------------------------------------------------------------------------------------------------
2023-10-23 14:53:23,750 epoch 8 - iter 27/275 - loss 0.04936288 - time (sec): 1.46 - samples/sec: 1528.40 - lr: 0.000010 - momentum: 0.000000
2023-10-23 14:53:25,133 epoch 8 - iter 54/275 - loss 0.04354247 - time (sec): 2.84 - samples/sec: 1603.51 - lr: 0.000009 - momentum: 0.000000
2023-10-23 14:53:26,515 epoch 8 - iter 81/275 - loss 0.03367946 - time (sec): 4.22 - samples/sec: 1558.37 - lr: 0.000009 - momentum: 0.000000
2023-10-23 14:53:27,889 epoch 8 - iter 108/275 - loss 0.02870090 - time (sec): 5.60 - samples/sec: 1632.80 - lr: 0.000009 - momentum: 0.000000
2023-10-23 14:53:29,313 epoch 8 - iter 135/275 - loss 0.02468517 - time (sec): 7.02 - samples/sec: 1619.99 - lr: 0.000008 - momentum: 0.000000
2023-10-23 14:53:30,564 epoch 8 - iter 162/275 - loss 0.02304861 - time (sec): 8.27 - samples/sec: 1652.63 - lr: 0.000008 - momentum: 0.000000
2023-10-23 14:53:31,821 epoch 8 - iter 189/275 - loss 0.02146790 - time (sec): 9.53 - samples/sec: 1655.76 - lr: 0.000008 - momentum: 0.000000
2023-10-23 14:53:33,087 epoch 8 - iter 216/275 - loss 0.01984298 - time (sec): 10.80 - samples/sec: 1675.71 - lr: 0.000007 - momentum: 0.000000
2023-10-23 14:53:34,347 epoch 8 - iter 243/275 - loss 0.01897873 - time (sec): 12.06 - samples/sec: 1676.08 - lr: 0.000007 - momentum: 0.000000
2023-10-23 14:53:35,613 epoch 8 - iter 270/275 - loss 0.01918661 - time (sec): 13.32 - samples/sec: 1673.72 - lr: 0.000007 - momentum: 0.000000
2023-10-23 14:53:35,847 ----------------------------------------------------------------------------------------------------
2023-10-23 14:53:35,848 EPOCH 8 done: loss 0.0188 - lr: 0.000007
2023-10-23 14:53:36,372 DEV : loss 0.16693733632564545 - f1-score (micro avg) 0.8722
2023-10-23 14:53:36,377 ----------------------------------------------------------------------------------------------------
2023-10-23 14:53:37,694 epoch 9 - iter 27/275 - loss 0.01861865 - time (sec): 1.32 - samples/sec: 1575.51 - lr: 0.000006 - momentum: 0.000000
2023-10-23 14:53:38,944 epoch 9 - iter 54/275 - loss 0.01361342 - time (sec): 2.57 - samples/sec: 1738.65 - lr: 0.000006 - momentum: 0.000000
2023-10-23 14:53:40,188 epoch 9 - iter 81/275 - loss 0.01073145 - time (sec): 3.81 - samples/sec: 1744.90 - lr: 0.000006 - momentum: 0.000000
2023-10-23 14:53:41,460 epoch 9 - iter 108/275 - loss 0.01121247 - time (sec): 5.08 - samples/sec: 1751.86 - lr: 0.000005 - momentum: 0.000000
2023-10-23 14:53:42,707 epoch 9 - iter 135/275 - loss 0.01189528 - time (sec): 6.33 - samples/sec: 1781.63 - lr: 0.000005 - momentum: 0.000000
2023-10-23 14:53:44,008 epoch 9 - iter 162/275 - loss 0.01615819 - time (sec): 7.63 - samples/sec: 1802.89 - lr: 0.000005 - momentum: 0.000000
2023-10-23 14:53:45,277 epoch 9 - iter 189/275 - loss 0.01994894 - time (sec): 8.90 - samples/sec: 1780.91 - lr: 0.000004 - momentum: 0.000000
2023-10-23 14:53:46,516 epoch 9 - iter 216/275 - loss 0.01836120 - time (sec): 10.14 - samples/sec: 1753.83 - lr: 0.000004 - momentum: 0.000000
2023-10-23 14:53:47,802 epoch 9 - iter 243/275 - loss 0.01720199 - time (sec): 11.42 - samples/sec: 1749.97 - lr: 0.000004 - momentum: 0.000000
2023-10-23 14:53:49,092 epoch 9 - iter 270/275 - loss 0.01601844 - time (sec): 12.71 - samples/sec: 1770.01 - lr: 0.000003 - momentum: 0.000000
2023-10-23 14:53:49,320 ----------------------------------------------------------------------------------------------------
2023-10-23 14:53:49,320 EPOCH 9 done: loss 0.0158 - lr: 0.000003
2023-10-23 14:53:49,846 DEV : loss 0.17052848637104034 - f1-score (micro avg) 0.8771
2023-10-23 14:53:49,851 saving best model
2023-10-23 14:53:50,424 ----------------------------------------------------------------------------------------------------
2023-10-23 14:53:51,862 epoch 10 - iter 27/275 - loss 0.00049426 - time (sec): 1.44 - samples/sec: 1504.14 - lr: 0.000003 - momentum: 0.000000
2023-10-23 14:53:53,265 epoch 10 - iter 54/275 - loss 0.00551470 - time (sec): 2.84 - samples/sec: 1553.44 - lr: 0.000003 - momentum: 0.000000
2023-10-23 14:53:54,689 epoch 10 - iter 81/275 - loss 0.00500161 - time (sec): 4.26 - samples/sec: 1574.51 - lr: 0.000002 - momentum: 0.000000
2023-10-23 14:53:56,024 epoch 10 - iter 108/275 - loss 0.00422058 - time (sec): 5.60 - samples/sec: 1559.22 - lr: 0.000002 - momentum: 0.000000
2023-10-23 14:53:57,322 epoch 10 - iter 135/275 - loss 0.00602318 - time (sec): 6.90 - samples/sec: 1566.35 - lr: 0.000002 - momentum: 0.000000
2023-10-23 14:53:58,708 epoch 10 - iter 162/275 - loss 0.00524299 - time (sec): 8.28 - samples/sec: 1607.88 - lr: 0.000001 - momentum: 0.000000
2023-10-23 14:53:59,960 epoch 10 - iter 189/275 - loss 0.00665176 - time (sec): 9.53 - samples/sec: 1624.64 - lr: 0.000001 - momentum: 0.000000
2023-10-23 14:54:01,208 epoch 10 - iter 216/275 - loss 0.00581924 - time (sec): 10.78 - samples/sec: 1646.89 - lr: 0.000001 - momentum: 0.000000
2023-10-23 14:54:02,468 epoch 10 - iter 243/275 - loss 0.00887703 - time (sec): 12.04 - samples/sec: 1682.46 - lr: 0.000000 - momentum: 0.000000
2023-10-23 14:54:03,751 epoch 10 - iter 270/275 - loss 0.00927128 - time (sec): 13.33 - samples/sec: 1683.37 - lr: 0.000000 - momentum: 0.000000
2023-10-23 14:54:03,981 ----------------------------------------------------------------------------------------------------
2023-10-23 14:54:03,981 EPOCH 10 done: loss 0.0107 - lr: 0.000000
2023-10-23 14:54:04,511 DEV : loss 0.1703922599554062 - f1-score (micro avg) 0.8806
2023-10-23 14:54:04,517 saving best model
2023-10-23 14:54:05,535 ----------------------------------------------------------------------------------------------------
2023-10-23 14:54:05,536 Loading model from best epoch ...
2023-10-23 14:54:07,515 SequenceTagger predicts: Dictionary with 25 tags: O, S-scope, B-scope, E-scope, I-scope, S-pers, B-pers, E-pers, I-pers, S-work, B-work, E-work, I-work, S-loc, B-loc, E-loc, I-loc, S-object, B-object, E-object, I-object, S-date, B-date, E-date, I-date
2023-10-23 14:54:08,212
Results:
- F-score (micro) 0.9069
- F-score (macro) 0.7836
- Accuracy 0.8398
By class:
precision recall f1-score support
scope 0.8971 0.8920 0.8946 176
pers 0.9760 0.9531 0.9644 128
work 0.8533 0.8649 0.8591 74
object 0.6667 1.0000 0.8000 2
loc 0.3333 0.5000 0.4000 2
micro avg 0.9081 0.9058 0.9069 382
macro avg 0.7453 0.8420 0.7836 382
weighted avg 0.9109 0.9058 0.9080 382
2023-10-23 14:54:08,213 ----------------------------------------------------------------------------------------------------
|